Critique of pure folly.
Proposals to discard the current systems of biological nomenclature and replace them with a so-called phylogenetic taxonomy received considerable publicity for about a decade without any specific details on implementation of the replacement system (see Nixon & Carpenter, 2000). This lack of detail did not prevent--indeed, it doubtless aided--the favorable fanfare in the popular media (e.g., Pennisi, 1996, 2001; Chui, 1999; Milius, 1999; Fellman, 2000; Withgott, 2000). Finally, a draft of a "PhyloCode" has been promulgated on the World Wide Web at <http:/Iwww.ohiou.edu/phylocode/>. The release of the draft presents the opportunity, as the preface states, "to explore the ramifications of phylogenetic [sic] nomenclature." The philosophical "foundation" of the PhyloCode has already been the subject of a thorough critique by Nixon and Carpenter (2000), who also showed that so-called phylogenetic nomenclature would be less stable than the existing systems. The response by advocates of the PhyloCode to that critique h as been mere plays on words (viz., Lee, 2001; Bryant & Cantino, 2002); substantive counterargument is not likely to be forthcoming. Rather than a dissection of these specious responses here, the availability of the draft PhyloCode invites another kind of critique, a scrutiny of the details of this attempt to provide "a precise, coherent, international system for naming" (Cantino & de Queiroz, 2000: preamble 1), which will obviate the necessity for any such rebuttal.
Some background will be useful. Since the inception of so-called phylogenetic taxonomy (de Queiroz & Gauthier, 1990, etc.), its advocates have claimed that the advantage of this approach would be greater taxonomic stability than current systems of nomenclature provide. Whereas the current Linnaean systems employ typification (to family-group level) and ranked taxa, the proposed "node-based" method of so-called phylogenetic taxonomy entails designating two (or more) terminals subtended by a node of a cladogram and declaring that the name points to the common ancestor of those descendants and all other descendants of that ancestor. Arguments purporting to establish that this approach is more stable have emphasized changes in rank due to lumping and splitting in current nomenclature, and abolition of ranks has been proposed as a solution. I will return to the subject of ranks later, but here wish to stress that the crucial aspect of taxonomic stability is content. As developed by Nixon and Carpenter (2000), taxo nomic stability is measured in terms of all taxa included in a group referred to by a particular name; or, put another way, instability is measured by the number of included taxa that do not occur in each application of a particular name. A name is used in two different senses if it does not refer to the same set of taxa in two applications; that is instability, in taxonomy and in language. By contrast, if a name is used in one case, say, for a family, and another for a subfamily, but in each case circumscribes exactly the same included taxa, that is trivial for communicating the meaning--the content--of the name. Any experienced taxonomist understands this, but advocates of the PhyloCode completely fail to grasp this point. For example, Lee (1996: 188) approved of the fact that in so-called phylogenetic taxonomy "it is not the names of clades that change, but the contents of those named clades"; and Bryant and Cantino (2002: 53) concluded that the PhyloCode achieved "stability in the phylogenetic meaning of taxon names" despite conceding instability in terms of content, although they allowed that that was "a relevant issue." This confusion between meaning and spelling persists in the draft PhyloCode. In fact, the stability claimed for the PhyloCode is purely metaphysical (Nixon & Carpenter, 2000), having nothing to do with real taxonomic stability; by Orwellian antilogic, its instability is proclaimed to be "stable."
Net change in content quantifies taxonomic instability: If two uses of a name circumscribe the same taxa, the net change is zero, and there is no instability. Nixon and Carpenter (2000) presented examples showing that Linnaean classification outperforms or matches the stability of so-called phylogenetic taxonomy and that if a Linnaean classification is based on a type that is one of the "specifiers" of a "node-based" PhyloCode classification, it is always possible to maintain a classification at least as stable as, and usually more stable, than the PhyloCode classification.
Measuring net change in content also provides a basis for maximizing stability of established names when change is required, as for example when revising classification to remove paraphyly. This can be illustrated with an example, taken from a proponent of the PhyloCode, in which the consequences of nomenclatural changes in the so-called phylogenetic and Linnaean system were compared--superficially. De Queiroz (1996: figs. 6, 8) presented a hypothetical example comparing changes in composition of taxa under both Linnaean and so-called phylogenetic classification to show that "both Linnaean and phylogenetic approaches can result in changes in the hypothesized composition of taxa, that is, instability" (p. 15). Indeed, but it is generally less under the Linnaean system. The example is reprised in Figures 1 and 2, which show two different cladograms for the same group. The so-called phylogenetic and the Linnaean classifications presented by de Queiroz (1996) for each cladogram are given in Tables I and II.
In this case the taxon changing position, Alpha, is a specifier (or pseudotype) in the so-called phylogenetic classification, and a type in the Linnaean classification. The consequences of its shift in position for the so-called phylogenetic classification are that Alphadeltina goes from encompassing four taxa to eight (4), Alphathetonia goes from encompassing eight taxa to five (3), and Thetazina undergoes no change (0), for a net total of seven (7) changes in content. For the Linnaean classification, de Queiroz (1996: 14) noted: "In this case, there are several possible nomenclatural outcomes, depending on which taxa are assigned to the family category. Let us just consider the ranking scheme that most closely approximates the situation that I described earlier for the phylogenetic [sic] definitions[;] ... that is, the one in which all eight terminal taxa are considered to form a single family." In that case the family Alphaceae replaces the suborder Alphineae and goes from referring to four taxa to eight ( 4), whereas Thetaceae does not change in composition, but merely in rank, to Thetoideae (0), for a net total of four (4) changes in content. Not only are the changes fewer, the change in rank endings actually conveys information on relative position in the hierarchy. I will discuss the subject of ranks below, but first wish to stress that the changes for the so-called phylogenetic classification are mandatory: Under a "node-based" system, when the named clade pointed to by one of the specifiers of that clade is discarded in favor of a differently circumscribed clade, whatever clade now pointed to by both specifiers adopts the name. The Linaaean system is more flexible, because only the taxon containing the single type must retain the name established with that type. So, for this example one of the possible nomenclatural outcomes not discussed by de Queiroz would be to retain the suborder Alphineae, retain the family Thetaceae, and restrict the family Alphaceae to Alpha (3), while creating a new family for Bet a, Delta, and Gamma--a net change of 3 for the previously established names. An experienced taxonomist might well prefer this outcome, depending on historical usage, etc.; it also might be deemed preferable due to fewer changes in the established names.
This example was repeated in de Queiroz (1997) and also adapted for the popular media (Withgott, 2000: 649), where the changes in the Linnaean system similar to those proposed by de Queiroz were characterized as "the most likely," and thus, "[u]nder most circumstances, PhyloCode proponents say, the Linnaean system experiences more name changes due to the necessity of fitting groups into categorical ranks and applying appropriate suffixes." As shown above, that misrepresents the actual situation; it is the Linnaean system that generally can undergo fewer changes in group content, and that its flexibility allows this property--stability--to be maximized.
Returning to the subject of ranks, advocates of the PhyloCode have ignored the actual measure of taxonomic stability in terms of content, instead emphasizing changes in rank in Linnaean classification and thus instability due to categorical assignment. Examples presented to the popular press (e.g., Kron, 1997; see also Withgott, 2000: 649) purport to show that parallel construction of Linnaean and so-called phylogenetic classifications demonstrate the latter to "result in the fewest name changes." Content is not measured in such examples; only rank changes are counted. Of course, advocates of the PhyloCode change ranks too, by doing away with ranks entirely. This is portrayed as an advantage, but the only stability thereby attained is in orthography, and much is discarded (Nixon & Carpenter, 2000). Any hierarchy is inherently ranked, whether a cladogram or formal Linnaean classification. The named categories in a Linnaean classification convey information on exclusivity: Taxa at the same rank are not included within one another; taxa at higher ranks are not contained within taxa of lower ranks. This information is immediately communicated in a given case simply by stating the ranks or, even more efficiently, by the use of standardized suffixes for family-group categories in the classification and binominal nomenclature for species. The suffixes are relatively few in number, and readily learned, as is the convention of binomina, even by nonspecialists or the general public. In contrast, in so-called phylogenetic rankless classifications no indication of relationships of taxa is communicated by the names at all. Reference to a cladogram is required to communicate any information on group membership, thus a so-called phylogenetic classification would be understood and used only by specialists. And yet proponents of the PhyloCode claim--ironically--that communicating nothing about relationships would be easier, because:
"Suppose I have discovered a clade--let's just call it Bob." ... As modern evolutionary information piles in, Bob might turn out to include a huge number of species from another part of the tree, or it might get transferred to the midst of what had been a different clade. Since the name Bob did not connote any rank, it could stay the same, as could the names of its new relatives. "I can do all this [with Bob], and it doesn't have any effect on Ted, Carol, or Alice," Donoghue explains. (Milius, 1999: 268)
Advocates of the PhyloCode even claim that the loss of information conveyed by ranks, and the cumbersome necessity of reference to a cladogram, would promote classificatory activity: "People are scared to name something--they're afraid of the nomenclatural consequences. ... It's happened to me dozens of times where I've discovered a major new clade of importance and [didn't name it] because of bad downstream consequences" (Donoghue, quoted in Withgott, 2000: 650). Despite such claims, however, the drafters of the PhyloCode have found exclusion information desirable and have installed a cumbersome approximation of ranks in the PhyloCode, as will be seen.
IV. The PhyloCode
The draft PhyloCode is dated April 8, 2000; the latest revision date on the Web page where it was posted, <http://www.ohiou.edu/phylocode/>, was July 1, 2002, at the time this issue of the Botanical Review went to press. It is available in several formats, and the drafters admit that it is incomplete, because it "governs only the naming of clades. Rules governing species names will be added to a later version." The incompleteness is due to the fact that the drafters cannot agree on what form species names should take (see Cantino et al., 1999). Thus, from the outset, the PhyloCode is not universal. The situation is even worse: The draft is "designed so that it may be used concurrently with the existing codes based on Linnaean nomenclature"! Obviously, adopting any independent parallel system to those currently in use will only create nomenclatural chaos, not stability, and this would be exacerbated by a hybrid system using Linnaean species nomenclature but PhyloCode names for higher taxa. This hybrid is promo ted despite the fact that "[t]he differences between phylogenetic [sic] and rank-based nomenclature are just too fundamental for them to be combined" (Cantino, quoted in Pennisi, 2001: 2304).
Of course, it is a canard to claim that the Linnaean system cannot convey phylogenetic relationships; that it can was emphasized, for example, by Hennig (1966). Linnaean classification and cladograms are both hierarchical, and therefore any cladogram can be described in a Linnaean hierarchy, limited only by the number of formal ranks and by whether some nodes are not named, are given informal names (which the current codes do not forbid), or are identified using a convention such as phyletic sequencing (see Wiley, 1979). But the self-contradiction implicit in promotion of the hybrid is thought to be necessary because "[o]ne can never know for sure that one system is better than another until both have been tried for a while" (Cantino, quoted in Pennisi, 2001: 2304).
That is plainly not true. Various aspects of so-called phylogenetic taxonomy have already been shown to be defective compared with Linnaean taxonomy (Nixon & Carpenter, 2000), in both practical and philosophical terms. This can also be shown by consideration of the details of the draft PhyloCode, which follows.
The preface to the draft PhyloCode states that "clades and species require names that explicitly and unambiguously refer to those entities and do not change with time." The preface then goes on to assert that the present codes do not provide such names and cites name changes consequent to new combinations (species) or changes in rank (clades). That is, stability is discussed only with reference to reassignments; as is usual with proponents of so-called phylogenetic taxonomy, the crucial aspect of taxonomic stability, content (meaning) is not even mentioned. That failure is significant, for consideration of it shows the Linnaean systems to be superior (Nixon & Carpenter, 2000). Thus, the preface states: "a supraspecific name is given a phylogenetic definition and applies to whatever clade fits that definition, regardless of its hypothesized content." The PhyloCode is therefore to provide the opposite of unambiguous names, for a given name can vary wildly in content (see the paleoherb example in Nixon & Carpent er, 2000, all parts of which example are completely valid under the PhyloCode, refuting the claim in Bryant & Cantino, 2002, that they are misleading), but no indication of this can be made. And the situation will be worse, because the PhyloCode is proposed for concurrent use with Linnaean nomenclature (and worse still, in hybrid form, because species names are not governed by the draft PhyloCode). Moreover, although "rules and recommendations have been included to ensure that most names will be used in ways that approximate their current usage" [if this is desirable, then why is the PhyloCode necessary?], "taxon membership associated with a name will sometimes differ somewhat under the Linnaean and phylogenetic [sic] systems," so a forthcoming companion volume "will provide phylogenetic [sic] definitions for many widely used dade names." That is to say, taxa will be renamed, in terms of content (or meaning), as ideas on phylogeny change, which is hardly a prescription for stability, either during or after an y transition period.
The advantages claimed for the PhylaCode in the preface are concerned mostly with improved nomenclatural stability--a false claim, as previously shown. But the last claimed benefit is that abandonment of ranks will eliminate subjectivity in taxonomy, due to the arbitrary nature of ranks. But it is also stated earlier: "Another novel aspect of the PhyloCode is that it permits the taxonomist to restrict the application of names with respect to dade composition. If a taxonomist wishes to ensure that a name refers to a dade that either includes or excludes particular subtaxa, the definition may contain a qualifying clause specifying conditions under which the name cannot be used."
Information on exclusion is something conveyed automatically by ranks, of course. It seems that ranks have desirable properties after all. I will examine the qualifying clauses below, but this will not be the last unintentional irony uncovered in the PhyloCode.
The preamble begins by stating that: "Biology requires a precise, coherent, international system for naming clades and species of organisms. The PhyloCode attempts to satisfy this need by providing rules for naming clades (this version) and species (future versions) and describing the nomenclatural principles that form the basis for those rules." The draft PhyloCode does not provide rules for naming species, thus it certainly does not satisfy the stated need, nevertheless, it "may be used concurrently with the preexisting codes." The recipe for nomenclatural chaos is even worse, for "Although the PhyloCode relies on the preexisting codes... to determine the acceptability of preexisting names, it governs the application of those names independently from the preexisting codes." (For discussion of the precision and coherence of so-called phylogenetic taxonomy, see Nixon & Carpenter, 2000.)
The six principles of the draft PhyloCode are rife with contradiction. Taken seriatim:
"1. Reference. The primary purpose of taxon names is to provide a means of referring to taxa, as opposed to indicating their characters, relationships, or membership." As if taxa can be referred to without some indication of attributes or membership, let alone the fact that for clades to be specified there has to be some indication of relationships.
"2. Clarity. Taxon names should be unambiguous in their designation of particular taxa. Nomenclatural clarity is achieved through explicit definitions." It is scarcely unambiguous if a name does not change but its meaning (content, or "designation of particular taxa") does.
"3. Uniqueness. To promote clarity, each taxon should have only one accepted name, and each accepted name should refer to only one taxon." Names can scarcely be unique if they can apply to taxa differing greatly in content.
"4. Stability. The names of taxa should not change over time. As a corollary, it must be possible to name newly discovered taxa without changing the names of previously discovered taxa." It is scarcely stable if names do not "change" while at the same time changing greatly in content.
"5. Phylogenetic context. The PhyloCode is concerned with the naming of taxa and the application of taxon names within a phylogenetic context." But the first principle stated that taxa are to be referred to without indication of relationships.
"6. The PhyloCode permits freedom of taxonomic opinion with regard to hypotheses about relationships; it only concerns how names are to be applied within the context of a given phylogenetic hypothesis." True, but it is precisely how those names are applied that is the source of the instability of the PhyloCode names relative to the Linnaean system (see above, and Nixon & Carpenter, 2000).
I will not discuss all 21 articles of the draft PhyloCode, many of which are modeled on the current Linnaean codes, as well as the (failed) draft BioCode (Greuter et al., 1998). Some of the problems noted above in preceding sections could be flogged again, but, sparing the reader, I shall limit discussion to a few specific articles.
1. Article 6
This article is concerned with the status of names, whether established under the PhyloCode or one of the current codes. It contains the following recommendation (which is not mandatory, in contrast to rules): "Recommendation 6. 1B. In order to indicate which names are established under this code and therefore have explicit phylogenetic definitions (and whose endings are not reflective of rank), it may be desirable to distinguish these names from supraspecific names governed by preexisting codes, particularly when both are used in the same publication."
Two examples are then presented, the first of which, adding the letter "P" for PhyloCode names and "L" for Linnaean names, simply deserves the appellation "clumsy." The second is more interesting: applying the term "clade" to a PhyloCode name and the categorical rank to a Linnaean name when they apply to the same taxon, e.g., Glade Teucrioideae versus Subfamily Teucrioideae. That, of course, is a distinction that could be applied in the absence of a PhyloCode and already is applied in the use of informal supraspecific names in current taxonomy (typically ranked). Because, of course, not all clades are named, either in current taxonomy or in the draft PhyloCode (Article 2), perhaps widespread adoption of this as an informal (even rankless) category would be a sufficient and harmless sop for those who are too busy discovering clades to bother with naming them according to current nomenclatural rules.
2. Article 9
This article is concerned with the general requirements for establishing clade names. It clearly states that knowledge of phylogeny is a requirement: "9.4. In order to be established, a clade name must be provided with a phylogenetic definition, written in English or Latin, linking it explicitly with a particular clade. The name applies to whatever clade fits the definition. Shortly thereafter appears a clarification: "Note 9.4.2: The application of a phylogenetic definition, and thus also a phylogenetically defined clade name, requires an hypothesized phylogeny."
Again the first principle of the PhyloCode is contradicted, but that aside, this emphasizes again the need for a cladogram to accompany names, to communicate anything about what is included or excluded. And the hypothesized phylogeny is a prerequisite; it must exist before a name can be proposed. However, "Recommendation 9B. Establishment of names for poorly supported clades should be done with careful consideration of possible nomenclatural consequences if the phylogenetic hypothesis turns out to be incorrect. It may frequently be advisable to use only informal names for poorly supported clades."
Support, poor or otherwise, is not defined, but anyway, this is merely a recommendation, because, after all, not having to be concerned with "downstream consequences" is one of the motivations for the PhyloCode, so clades in fact can be named that are completely unsupported by character evidence. Finally, "Recommendation 9E. In order to facilitate referral of species that are not cited in the definition of a clade name or included in the phylogenetic analyses that provide evidence for the existence of the clade, it will often be worthwhile to include in the protologue an English or Latin description or diagnosis, a list of synapomorphies, and/or a list of included taxa."
Protologue, according to the PhyloCode's glossary, is "for example, description, diagnosis, phylogenetic definition, registration number, designation of type, illustrations, references, synonymy, geographical data, specimen citations, and discussion." Yes, it would certainly be worthwhile to include some such information when establishing a name, but it is not necessary; this is a recommendation. It is not necessary to study the characters, or even look at specimens (Nixon & Carpenter, 2000), as is confirmed by this recommendation: "Recommendation 9C. Conversion of preexisting names to clade names should be done with a thorough knowledge of the group concerned, including its taxonomic and nomenclatural history and previously used diagnostic features. Wholesale conversion of preexisting names by authors who have not worked on the systematics of the groups concerned is strongly discouraged."
3. Article 11
This article is concerned with specifiers and qualifying clauses, and is one of the most revealing articles. Specifiers are "species, specimens, or synapomorphies"; Nixon and Carpenter (2000) termed these "pseudotypes." These specifiers must include "species or specimens," either entirely or in part (for "apomorphy-based" definitions). Because species names are not governed by the PhyloCode, the hybrid nature of the proposed system is again highlighted. This is further emphasized in: "11.5. Specimens that are not types may be used as specifiers only if the specimen does not belong to a named species under any code."
The dependence of the PhyloCode on current codes actually goes even farther: "11.8. In the interest of consistency with the preexisting codes, a clade whose name is converted from a genus name under a preexisting code, or is derived from the stem of a genus name, should include the type of the genus name. Therefore, when a dade name is converted from a preexisting genus name or is a new or converted name derived from the stem of a genus name, the definition of the dade name must use the type species of that genus as an internal specifier."
And relation to current codes becomes ironic with: "11.9. In order to restrict the application of a name with respect to dade composition (i.e., under alternative hypotheses of relationship), phylogenetic definitions may include qualifying clauses specifying conditions under which the name cannot be applied to any dade (see Example 1). It is also possible to restrict dade composition under alternative hypotheses of relationship through careful wording of definitions (see Examples 2 and 3)."
The qualifying clauses turn out to be a cumbersome approximation of ranks. Taking the examples in turn:
Example 1: The name Pinnipedia is traditionally applied to a group composed of sea lions (Oatriiidae), walruses (Odobenidae), and seals (Phocidae). However, under some phylogenetic hypotheses, the sister group of one or more of these taxa is a group of terrestrial carnivorans. If the name Pinnipedia is defined as "the dade stemming from the most recent common ancestor of Otaria byronia de Blainville 1820, Odobenus rosmarus Linnaeus 1758, and Phoca vitulina Linnaeus 1758, provided that it possessed flippers homologous with those in the aforementioned species," then the name would not be applicable to any dade in the context of phylogenetic hypotheses in which the most recent common ancestor of these species was inferred not to have had flippers. The phrase "provided that it possessed flippers homologous with those in the aforementioned species" is a qualifying clause.
Ordinal names are not regulated in the International Code of Zoological Nomenclature, but in ranked Linnaean classification Pinnipedia is an order. As such, it is not contained within the order Carnivora, and vice versa. If in fact the support for a sister-group relationship between part of Pinnipedia and part of Carnivora is deemed sufficiently strong [oops--sorry, that doesn't matter], so that reclassification is required to reflect changed ideas on phylogeny [oops, sorry again-I forgot that Linnaean classification can't reflect phylogeny], then the sense in which each ordinal name is used would have to change. This could be done in a variety of ways, depending on the specific details, for example, through lumping of two orders into one, or recognizing more than two orders. It would be possible to measure the net change in content for each established name and to choose among courses that minimize the instability. The exclusivity of ranks is automatic, but the PhyloCode cannot convey this information by the name and so must resort to exclusion with a verbose qualifying phrase to limit the application of the name.
Example 2 is a lengthy exposition on the possibilities for three taxa with relationships at first depicted as: (Crocodylus niloticus (Lacerta agilis Youngina capensis)). Several higher names are defined using "stem-based" definitions, as follows: Lepidosauriformes referring to the most inclusive clade containing L. agilis but not Y capensis, Lepidosauromorpha referring to the most inclusive clade containing L. agilis but not C. niloticus, and Archosauromorpha referring to the most inclusive clade containing C. niloticus but not L. agilis. Another name is "node based": Sauna, the clade stemming from the most recent common ancestor of L. agilis and C. niloticus.
The possibility that the cladogram turns out to be: (Y. capensis (C. niloticus L. agilis)) has as consequence that Lepidosauromorpha is now nested within Lepidosauriformes, reversing the former hierarchical relationships of the names. (Also, Y. capensis is no longer in Sauna, but never mind.) Thus:
In order to prevent these names from reversing their hierarchical relationships, the name Lepidosauriformes could be defined as "the most inclusive subclade of Sauria (Clade (Lacerta agilis + Crocodylus niloticus)) containing Lacerta agilis but not Youngina capensis," in which case Lepidosauriformes would become a synonym of Lepidosauromorpha (rather than the name of a more inclusive clade) in the context of the new phylogenetic hypothesis.... The first definition leaves the application of the name Lepidosauriformes unrestricted; the second definition restricts its application to a subclade of Sauria. However, the same restriction can be achieved by defining Lepidosauriformes as the most inclusive clade containing Lacerta agilis but not Youngina capensis or Crocodylus niloticus.
The same restriction is achieved, quite easily, through use of appropriate ranks (actually, because priority does not apply in zoology above the level of family-group, the name for Lepidosauromorpha = Lepidosauriformes could be decided on whatever basis deemed appropriate. This example really illustrates nothing more than the inability of so-called phylogenetic nomenclature to retain the intended meaning of a name as ideas on phylogeny change; see Nixon & Carpenter, 2000).
The third example could also be dealt with as intended by ranked taxa, but it would require specification of types and another name to illustrate the point:
Example 3: If a name is defined through a stem-based definition with more than one internal specifier, and one internal specifier is later found to be more closely related to the external specifier than to the other internal specifier, the definition does not apply to any clade. For example, suppose the name Halecostomi had been defined as referring to the most inclusive clade comprising Amia calva Linnaeus 1766 and Perca fluviatilis Linnaeus 1758 but not Lepisosteus osseus Linnaeus 1758. And suppose that subsequent analyses indicated that Lepisosteus osseus and Perca fluviatilis share a more recent common ancestor with one another than either does with Amia calva. If so, then there is no clade that fits the definition of Halecostomi (because there is no clade that includes both Amia calva and Perca fluviatilis but not Lepisosteus osseus), and that name cannot be used in the context of the accepted phylogeny.
In zoology, names above family-group level are not typified, but the same exclusivity could be conveyed by use of a taxon at the same rank as Halecostomi for Lepisosteus osseus.
Finally, even more ironically, although names in current use will change under the PhyloCode, stability is urged (albeit not required):
Recommendation 11A. Definitions of converted clade names should be stated in a way that attempts to capture the spirit of historical use to the degree that it is consistent with the contemporary concept of monophyly. Consequently, they should not necessitate, though they may allow, the inclusion of subtaxa that were historically excluded from the taxon. To accomplish this goal, internal specifiers of converted dade names should be chosen from among the set of taxa that were considered to form part of a taxon under either the original or traditional ideas about the composition of that taxon, and they should not include members of subtaxa that were not historically considered part of the taxon.
To illustrate this, an example is offered:
Example 1: The name Dinosauria was coined by Owen for the taxa Megalosaurus, Iguanodon, and Hylaeosaurus, and traditionally the taxon designated by that name has included these and certain other non-volant reptiles. It has not traditionally included birds. Although birds are now considered part of the dinosaur dade, the name Dinosauria should not be defined using any bird species as internal specifiers. Such a definition would force birds to be dinosaurs, thus trivializing the question of whether birds are dinosaurs. Instead, internal specifiers should be chosen from among taxa that have traditionally been considered dinosaurs; e.g., Megalosaurus bucklandi von Meyer 1832, Iguanodon anglicus Holl 1829, and Hylaeosaurus arm atus Mantell 1833.
How exactly the name Dinosauria would be specified so that it does not contain birds is not detailed; but, that aside, it is amusing that advocates of so-called phylogenetic nomenclature urge restriction of a name so that a notorious case of paraphyly not be indicated by altered usage of that name.
4. Article 15
Other articles concern themselves with matters such as priority ("precedence"), homonymy, synonymy, etc., as do current codes, including the conservation of names (Article 15). Conservation and rejection of names, obviously, is contrary to the notion that names "do not change with time" (preface) and the principles of uniqueness and stability.
5. Article 21
The final article of the draft PhyloCode is concerned with governance. This is to be vested in a "Society for Phylogenetic Nomenclature (SPN)," "an international, non-profit organization with no membership restrictions," about which it is stated: "Two committees of the SPN have responsibilities that pertain to this code: the International Committee on Phylogenetic Nomenclature (ICPN) and the Registration Committee." But "[t]hese organizations do not yet exist. They will be established before the PhyloCode is implemented." The SPN may not exist yet either, but perhaps it does, and perhaps it is synonymous with the drafters of the PhyloCode or the advisory group. Or perhaps it will be, if it is yet to be established. As matters currently stand with the draft, the drafters and advisory group are self-appointed, in contrast to the international representative bodies that govern the current codes of nomenclature. This explains why the recommendation embodied in the draft is for a concurrent, parallel system (albe it hybrid). It also explains the nearly exclusive self-citation in the preface. To be sure, published criticism of so-called phylogenetic taxonomy was less frequent during the 10 years prior to the draft PhyloCode than was advocacy, especially in the popular press, but this was doubtless because most taxonomists considered advocacy beneath notice, if not contempt. I had a similar attitude myself until recently (Nixon & Carpenter, 2000). Since then, the number of critiques of so-called phylogenetic taxonomy has burgeoned in the literature (e.g., Benton, 2000; Forey, 2001, 2002 and other articles in the February 2002 issue of the journal Taxon; Dyke, 2002; other articles in this issue of the Botanical Review), while the response by advocates has been anile (de Queiroz & Cantino, 2001; Bryant & Cantino, 2002). Mallet and Willmott (2003: 58) even refer to the PhyloCode as if it were already history: "One is reminded of the furore about 'rank-free taxonomy' and the PhyloCode... a few years ago." However, because o f the self-selection of advocates, the PhyloCode is unfortunately something that will not simply wither and die; it needs to be criticized by working taxonomists, in publications and in popular media, and condemned by international societies and organizations, especially those charged with regulating biological nomenclature.
The PhyloCode is proposed in the draft for coexistence with the current codes for some unspecified period of time. During that time, species names will continue to be regulated under the current codes. Therefore, it is acceptable that some authors use ranks, while others do not, or something in between. Acceptable definitions of PhyloCode names (Note 9.4.1) include "apomorphy-based" definitions with a single (Linnaean, if possible) type and character diagnosis, or node pointing, in the "node-based" and "stem-based" definitions, with multiple (possibly many) taxa as specifiers and no character diagnosis. And when species are eventually included, perhaps some authors will use binominals, while others will not (and so forth; see Cantino et al., 1999). Thus, the PhyloCode is basically a proposal to abandon one set of rules and enter a world of nomenclatural chaos.
The consequences for biological classification would obviously be disastrous, but as a language for scientific discourse the prospect is even more damning. Consider again that, under the PhyloCode, composition of taxa is unimportant, and change in composition is unimportant; the only thing that is important is that a taxon is "monophyletic" on any cladogram, even when it refers to something completely different in composition (meaning). Because the taxon is always monophyletic on any cladogram, it has no scientific value whatsoever. It is not a scientific hypothesis, it is merely a definition, in words that are divorced from reality. Proponents of the PhyloCode have offered as their primary criticism of the Linnaean system that it is "essentialistic"--as Nixon and Carpenter (2000) put it, how is the PhyloCode less essentialistic?
The PhyloCode should not be adopted, or even seriously considered, by anyone. It is pure folly.
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
Table I Node-pointing classifications presented by de Queiroz (1996: fig. 6) for the cladograms in Figures 1 and 2. Alphathetonia is defined as the taxon stemming from the most recent common ancestor of Alpha and Theta; Alphadeltina, as the taxon stemming from the most recent common ancestor of Alpha and Delta; and Thetazina, as the stemming from the most recent common ancestor of Theta and Zeta. Classification for Figure 1 Alphathetonia Alphadeltina Alpha Beta Delta Gamma Thetazina Epsilon Zeta Eta Theta Classification for Figure 2 Alphadeltina Beta Delta Gamma Alphathetonia Alpha Thetazina Epsilon Zeta Eta Theta Table II Linnaean classifications presented by de Queiroz (1996: fig. 8) for the cladograms in Figures 1 and 2. Alphineae is the suborder containing Alpha; Alphaceae is the family containing Alpha; Thetaceae is the family containing Theta; and Thetoideae is the subfamily containing Theta. Classification for Figure 1 Suborder Alphineae Family Alphaceae Alpha Beta Delta Gamma Family Thetaceae Epsilon Zeta Eta Theta Classification for Figure 2 Family Alphaceae = Family Thetaceae Beta Delta Gamma Alpha Subfamily Thetoideae Epsilon Zeta Eta Theta
I thank the working taxonomists who have begun to speak out against the follycode, among them Jerry Davis, who commissioned the manuscript, Jun-ichi Kojima, Toby Schuh, and Dennis Stevenson, for encouragement, and especially Kevin Nixon, who convinced me of the necessity of speaking out myself. Gareth Dyke provided corrections to the manuscript.
VII. Literature Cited
Benton, M. J. 2000. Stems, nodes, crown clades, and rank-free lists: Is Linnaeus dead? Biol. Rev. Cambridge Philos. Soc. 75: 633-648.
Bryant, H. N. & P. D. Cantino. 2002. A review of criticisms of phylogenetic nomenclature: Is taxonomic freedom the fundamental issue? Biol. Rev. Cambridge Philos. Soc. 77: 39-55.
Cantino, P. D. & K. de Queiroz. 2000. PhyloCode: A phylogenetic code of biological nomenclature. <http://www.ohiou.edu/phylocode/>.
-----, H. N. Bryant, K. de Queiroz, M. J. Donoghue, T. Eriksson, D. M. Hillis & M. S. Y. Lee. 1999. Species names in phylogenetic nomenclature. Syst. Biol. 48: 790-807.
Carpenter, J. M. & M. C. Day. 1988. Nomenclatural notes on Polistinac (Hymenoptera: Vespidae). Proc. Washington Entomol. Soc. 90: 323-328.
Chui, G. 1999. Class war means fight for survival. The Australian, September 29: 37.
De Queiroz, K. 1996. A phylogenetic approach to biological nomenclature as an alternative to the Linnacan systems in current use. In J. L. Reveal (ed.), Proceedings of a Mini-Symposium on Biological Nomenclature in the 21st Century, held at the University of Maryland on 4 November 1996. <http://www.inform.umd.edu/PBIO/nomcl/dequ.html>.
-----. 1997. The Linnacan hierarchy and the evolutionization of taxonomy, with emphasis on the problem of nomenclature. Aliso 15: 125-144.
----- & P. D. Cantino. 2001. Phylogenetic nomenclature and the PhyloCode. Bull. Zool. Nomencl. 58 (4): 254-271.
----- & J. Gauthier. 1990. Phylogeny as a central principle in taxonomy: Phylogenetic definitions of taxon names. Syst. Zool. 39: 307-322.
Dyke, G. J. 2002. Should paleontologists use "phylogenetic" nomenclature? J. Paleontol. 76: 793-796.
Fellman, B. 2000. What's in a name? Yale Alumni Mag., April: 36-39.
Forey, P. L. 2001. The PhyloCode: Description and commentary. Bull. Zool. Nomenel. 58 (2): 81-96.
-----. 2002. PhyloCode: Pain, no gain. Taxon 51: 43-54.
Greuter, W., D. L. Hawksworth, J. McNeill, M. A. Mayo, A. Minelli, P. H. A. Sneath, B. J. Tindall, P. Trehane & P. Tubbs. 1998. Draft BioCode (1997): The prospective international rules for the scientific names of organisms. Taxon 47: 127-150.
Hennig, W. 1966. Phylogenetic systematics. D. D. Davis & R. Zangerl, trans. Univ. of Illinois Press, Urbana.
Kron, K. A. 1997. Exploring alternative systems of classification. Aliso 15: 105-112.
Lee, M. S. V. 1996. The phylogenetic approach to biological taxonomy: Practical aspects. Zool. Scripta 25: 187-190.
-----. 2001. On recent arguments for phylogenetic nomenclature. Taxon 50: 175-180.
Mallet, J. & K. Willmott. 2003. Taxonomy: Renaissance or Tower of Babel? Trends Ecol. Evol. 18 (2): 57-59.
Milius, S. 1999. Should we junk Linnacus? Sci. News, October 23: 268.
Nixon, K. C. & J. M. Carpenter. 2000. On the other "Phylogenetic Systematics." Cladistics 16: 298-318.
Pennisi, E. 1996. Evolutionary and systematic biologists converge. Science 273: 181.
-----. 2001. Linnacus's last stand? Science 291: 2304.
Wiley, E. O. 1979. An annotated Linnaean hierarchy, with comments on natural taxa and competing systems. Syst. Zool. 28: 308-337.
Withgott, J. 2000. Is it "So long, Linnacus"? BioScience 50: 646-651.
|Printer friendly Cite/link Email Feedback|
|Author:||Carpenter, James M.|
|Publication:||The Botanical Review|
|Date:||Jan 1, 2003|
|Previous Article:||The Linnaean system and its 250-year persistence.|
|Next Article:||The illogical basis of phylogenetic nomenclature.|